The expected increase in energy demand in the United States has led to the pursuit of more efficient methods to generate thermal and electrical energy for the residential sector. One possible approach that could both increase generation efficiency and reduce CO₂ emissions is Combined Heat and Power (CHP). CHP plants, powered by natural gas, can act as an integrated residential utility supplier by producing the thermal and electrical energy needed to meet the heating, cooling, and electricity demands of a (future) residential neighborhood. However, a CHP plant operating in island (i.e., grid-disconnected) mode must be optimally sized to maximize efficiency and to lower the capital and marginal costs. In this paper, we describe a novel simultaneous optimization of design and operating strategies for a CHP plant as a utility producer for a residential neighborhood. The plant, operating in island mode, integrates distributed residential photovoltaic solar power generation, and is optimized to meet a time-dependent energy demand profile characteristic of residential energy use. To accurately capture the variability (hourly and seasonal) and uncertainty in residential energy demand and rooftop photovoltaic generation, a vast amount of energy data were incorporated in the problem formulation. The multi-scale optimization problem was solved using a temporal Lagrangean decomposition method, generating the design of a CHP plant that can efficiently meet all residential utility demands, taking into consideration the long-term (design) and short-term (operational) costs.

在美国，能源需求的预期增长导致人们寻求更有效的方法来为住宅部门产生热能和电能。一种可能既可以提高发电效率又可以减少CO _2的方法排放为热电联产（CHP）。由天然气驱动的热电联产厂可以通过生产满足（未来）住宅区的供热，制冷和电力需求所需的热能和电能，来充当综合住宅公用事业供应商。但是，必须优化以孤岛（即，电网断开）模式运行的热电联产工厂的规模，以最大程度地提高效率并降低资本和边际成本。在本文中，我们描述了CHP工厂作为居民区公用事业生产商的设计和运营策略的同时优化。该工厂以孤岛模式运行，集成了分布式住宅光伏太阳能发电，并经过优化以满足住宅能源使用随时间变化的能源需求特征。为了准确捕获住宅能源需求和屋顶光伏发电的变化性（每小时和季节性）和不确定性，在问题表述中纳入了大量的能源数据。考虑到长期（设计）和短期（运营）成本，使用时间拉格朗日分解法解决了多尺度优化问题，生成了可以有效满足所有住宅公用事业需求的热电联产厂的设计。